[chbot] Snubber circuits

[Trevor Wignall]

I have also had experience of resistors failing. In these applications, the
continuous power rating of the resistor is largely irrelevant. The
resistors also have a voltage rating, which may be 200V for some SMT parts
or perhaps 500V for smaller through-hole parts. The datasheets also give
pulsed power ratings, which can be difficult to interpret and relate but
for short irregular pulses, can be summarised as an energy rating. For 0805
SMT parts, this can be embarrassingly low - under 10mJ if I recall
correctly. Part of the problem with many of these resistors is that the
resistive element is a thin layer on the ceramic substrate, and for short
pulses, there isn't time to conduct the heat from this layer into the
substrate or to the end caps and the PCB, so it overheats while the rest of
the device is cool.
We found that when ESD (electrostatic discharge) testing, the low value
resistors were fine - the discharge current wasn't high enough to generate
enough voltage across the resistor to exceed its pulsed energy rating with
the available charge. The high value resistors were also fine - the voltage
across them was too high and found another path, such as across the
surface. But the mid-value resistors were hammered, with the voltage and
charge product exceeding their energy rating.
In a snubber circuit, the resistor may carry the peak load current as the
capacitor charges up after the contact opens. However when the contact
closes, the resistor will discharge the capacitor which may be charged up
to the peak supply voltage, so the energy being dissipated is up to 1/2 C
Vpeak^^2.
Note that the snubber will also have the supply voltage across it in one
state (depending on the configuration), so the capacitor will have pretty
much all this voltage across it and carry a current of 2 Pi F V C, which
will cause I^^2 R heating of the snubber resistor. (This doesn't apply for
DC of course, where F = 0.) Obviously the continuous power rating of the
snubber resistor should be much higher than this power.
Hope this helps
Trevor

On Wed, Nov 21, 2018 at 11:44 AM Peter Harris <petes.username at gmail.com>
wrote:

> Volker Kuhlmann wrote:
>
>> > It turned out that there was a spiral cut in the metal film at one end
>> of
>> > the resistor presumably to trim the value and when a voltage spike hit
>> the
>> > snubber it would spark across the film at the end of the spiral burning
>> > some film away.
>>
>> Very useful posting, thanks Peter! What was the voltage rating of the
>> resistors? And the spike voltage?
>>
>
> Sorry this happened in the mid 80's when data sheets were actual sheets of
> paper, I was a lowly technician and never saw the resistor specifications.
> We never tried to measure the voltage spike, we just left the test jig
> running with the new resistors until the boss was convinced the problem was
> solved.
>
>
>> Your experience would confirm 2 things if my suspicion is corect: The
>> snubber resistor must be rated for the spike voltage (the capacitor
>> probably too), and carbon film resistors are more tolerant to voltage
>> spikes exceeding their design rating. Or, resistant to spikes exceeding
>> their design wattage??
>>
>
> Under normal circumstances the capacitor will not see the full spike
> voltage, the resistor certainly will and must dissipate the energy in the
> spike.
> Something I learned with surface mount resistors is that a resistor's
> ability to dissipate heat is reduced when the energy is delivered in pulses
> that exceed its nominal rating. E.g. A 1W resistor array can take short
> pulses of several watts but the average power being dissipated under these
> conditions must be less than 1W. There are tables in the data sheets to
> calculate the required derating.
> I have no idea if similar derating has to be applied to through hole parts.
>
>   P
>
> _______________________________________________
> Chchrobotics mailing list Chchrobotics at lists.ourshack.com
> https://lists.ourshack.com/mailman/listinfo/chchrobotics
> Mail Archives: http://lists.ourshack.com/pipermail/chchrobotics/
> Meetings usually 3rd Monday each month. See http://kiwibots.org for
> venue, directions and dates.
> When replying, please edit your Subject line to reflect new subjects.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ourshack.com/pipermail/chchrobotics/attachments/20181121/5baef145/attachment-0001.html>